Module 2 - The Bias-Variance Tradeoff
Overview and Deliverables
This module introduces the bias-variance trade-off, which is one of the most fundamental ideas in machine learning and one which connects your data, the complexity of the problem you are solving, and the types of models which might be successful. We will introduce concepts such as in-sample and out-of-sample accuracy, generalization, overfitting, loss functions, and model assessment.
- 2/8: Submit Lab 1 on the Brightspace page for the course
- Start brainstorming project ideas and forming groups on slack.
Learning Objectives
- Loss functions and model assessment
- Overfitting and generalization error
- Bias-Variance Tradeoff
- Model Flexibility and Interpretability
- Basic Regression and Classification Models
- Introduction to scikit-learn package
Readings
- ISLP (Introduction to Statistical Learning): Sections 2.2-2.4
Optional:
- Chapter 2 of Hands on Machine Learning . I will go through an example of this in a coding vignette to introduce you to scikit-learn.